The Safe Maintenance Project
In 2000, the Dutch railway network consisted of about
6,500 railway kilometers, more than 3,000 switches,
and 400 yards. It handled about 5,000 trains per
day. In 2000, it transported over a million passengers
and 60,000 tons of freight per day. The network is
operated mainly by the Dutch railways (http://www.
ns.nl) and is fully financed and owned by the Dutch
government.
In 1995 in Mook, a train hit and killed three
track workers carrying out maintenance activities.
This accident caused a national investigation, which
showed that track workers had one of the most
dangerous jobs in The Netherlands. Their individual
risk (IR) was 3.4 deaths per 10,000 track workers
per annum, three times higher than the IR in the
building industry. The investigators concluded that
to ensure safety, the railways should allow no trains
on the parts of the rail infrastructure under maintenance.
The track workers should be protected in space
and time.
ProRail decided to combine maintenance activities
as much as possible and to change corrective maintenance
into preventive maintenance. In this way, it
would minimize the number of times maintenance
takes place and the total maintenance time, both of
which would improve safety and reduce the inconvenience
for train operators, passengers, and freight
transport.
ProRail divided the complete Dutch railway network
into working zones according to an unambiguous
method based on track functionality (den Hertog
et al. 2005). A working zone is the part of the railway
network taken out of service when track workers are
working on it. We call this practice protection in space.
After the working zones were in place, we developed
a maintenance schedule for the working zones,
which we call protection in time. This schedule was
based on a new concept for carrying out maintenance.
ProRail traditionally carried out maintenance
where and when necessary (corrective maintenance).